We've all seen films where highly-intelligent robots attempt to conquer Earth, with humans fighting to prevent their own extinction. These narratives usually end with human intelligence prevailing, or both species coexisting, the suggestion often being that human qualities will prevail over artificial intelligence (AI).
In reality, AI has gained considerable momentum in recent years, particularly with tools like ChatGPT. At the Olympic Games held in Paris this year, AI played a crucial role in the diving events by breaking down dives into sequences and analysing rotations, speeds, angles, and other factors in less than a tenth of a second.
AI typically requires substantial data input to properly analyse and generate content. By collecting diverse data, AI systems can enhance their knowledge on specific subjects with each analysis. This is often achieved by large corporations utilising user data to train their AI systems, as exemplified by companies like Microsoft.
In September, LinkedIn (owned by Microsoft) updated its privacy policy to permit the use of users’ personal data and content to train AI models for content generation (GenAI). Hong Kong's Privacy Commissioner for Personal Data (PCPD) expressed concerns about LinkedIn’s update, particularly since users' default consent was set to 'YES', causing alarm in various jurisdictions. As of 11 October 2024, LinkedIn has suspended the use of Hong Kong users’ data for GenAI training while addressing the PCPD’s concerns. Additionally, LinkedIn has updated its privacy policy to explain how user information is used to develop GenAI and has introduced an opt-out setting. Currently, LinkedIn does not train GenAI in Canada, the EU, EEA, UK, Switzerland, Hong Kong, and Mainland China.
This issue coincides with a joint statement issued on 29 October 2024 by the PCPD and 15 other global privacy authorities. This statement, sent to social media platforms and websites hosting publicly-accessible personal data (including Microsoft), updates an initial joint statement from August 2023. It offers further guidance on implementing dynamic, multi-layered approaches to protect against unlawful data scraping. The statement encourages entities to collaborate and develop solutions to address this privacy threat.
Key Takeaways
Although we are already deep into this new technology, there is currently no legislation in Hong Kong specifically governing AI. In June 2024, the PCPD published the "Artificial Intelligence: Model Personal Data Protection Framework", offering internationally recognised guidance to assist organisations in procuring, implementing, and using AI (including generative AI) while complying with the Personal Data (Privacy) Ordinance. This framework aims to help organisations harness AI benefits while safeguarding data privacy. Companies should carefully review this framework if they plan to introduce AI into their operations.
Another valuable resource is the EU's AI Act, the world’s first comprehensive AI regulation, which came into force on 1 August 2024. It is expected to serve as a template for jurisdictions planning similar legislation, alongside a landmark AI treaty designed to align AI systems with democratic values globally. It remains to be seen whether the Hong Kong Government will incorporate these provisions into future legislation - at present it is adopting a context-based approach rather than a comprehensive regime. It may prefer to look to the PRC model - the Mainland is in the process of rolling out its own targeted regulations on AI.
As the use of personal data for training AI models becomes more widespread, businesses and individuals will need to be increasingly mindful of the terms and conditions of the services they use, to understand how their data is being analysed and to make informed decisions about its use.
Given the speed with which AI is developing, those responsible for drafting laws to govern it are likely to struggle to deal with everything it is capable of - we are now in an era where science fiction has become reality. If, as seems likely, we will see machine intelligence that is also capable of truly independent thought, and even emotions, it will likely require an increasingly sophisticated legal framework. AI may acquire rights and protections, or we may shift towards a more restrictive regime, aimed at limiting AI from challenging our dominance as a species.
About Us
Howse Williams is a leading, full service, Hong Kong law firm. We combine the in-depth experience of our lawyers with a forward thinking approach.
Our key practice areas are corporate/commercial and corporate finance; commercial and maritime dispute resolution; clinical negligence and healthcare; insurance, personal injury and professional indemnity insurance; employment; family and matrimonial; trusts and wealth preservation; wills, probate and estate administration; property and building management; banking; fraud; distressed debt; investment funds; virtual assets; financial services/corporate regulatory and compliance.
As an independent law firm, we are able to minimise legal and commercial conflicts of interest and act for clients in every industry sector. The partners have spent the majority of their careers in Hong Kong and have a detailed understanding of international business and business in Asia.
Disclaimer: The information contained in this article is intended to be a general guide only and is not intended to provide legal advice. Please contact [email protected] if you have any questions about the article.